Use of Big Data to Support Regulatory Decision Making

Transcription

1 Use of Big Data to Support Regulatory Decision Making EMA-EuropaBio Annual Bilateral meeting Dr Alison Cave, Principal Scientific Administrator, Pharmacovigilance and Epidemiology Department An agency of the European Union

2 Objectives 1 Regulatory Challenges 1

3 2 An increasing number of medicines with genomic mechanism and/or genomic biomarkers. Smaller, focused RCTs which increases the importance of post-marketing monitoring.

4 An increasing number of medicines with genomic mechanism and/or genomic biomarkers. Smaller, focused RCTs which increases the importance of post-marketing monitoring. Innovation driving personalised prescribing creates regulatory challenges. 3

5 An increasing number of medicines with genomic mechanism and/or genomic biomarkers. Smaller, focused RCTs which increases the importance of post-marketing monitoring. Innovation driving personalised prescribing creates regulatory challenges. Increased activity in the rare disease area to meet unmet medical needs may be associated with more limited information at authorisation 4

6 5 An increasing number of medicines with genomic mechanism and/or genomic biomarkers. Smaller, focused RCTs which increases the importance of post-marketing monitoring. Innovation driving personalised prescribing creates regulatory challenges. Increased activity in the rare disease area to meet unmet medical needs may be associated with more limited information at authorisation Lack of information on generalisability of clinical trial results to normal clinical practice demands new approaches to gather complementary evidence

7 An increasing number of medicines with genomic mechanism and/or genomic biomarkers. Smaller, focused RCTs which increases the importance of post-marketing monitoring. Innovation driving personalised prescribing creates regulatory challenges. Increased activity in the rare disease area to meet unmet medical needs maybe associated with more limited information at authorisation Lack of information on generalisability of clinical trial results to normal clinical practice demands new approaches to gather complementary evidence Additional data sources are needed to better monitor risk/benefit in high risk groups often excluded from clinical trials 6

8 An increasing number of medicines with genomic mechanism and/or genomic biomarkers. Smaller, focused RCTs which increases the importance of post-marketing monitoring. Innovation driving personalised prescribing creates regulatory challenges. Welcome activity in the rare disease area to meet unmet medical needs is associated with more limited information at authorisation The high internal validity of clinical trials at the expense of external validity demands new approaches to gather complementary evidence Additional data sources are needed to appropriately monitor risk/benefit in high risk groups often excluded from clinical trials 7 Increasing interest in the development of combination therapies to treat complex diseases creates regulatory challenges

9 Objectives 1 2 Regulatory Challenges Which Big Data? 8

10 Which Big Data? Social media data/mhealth Clinical trial data Imaging data Omic data (Genetic, proteomic, metabolomic) Published literature Clinical databases (Prescriptions, EHRs and registries) Datasources Regulatory data (ADR, sales, Safety updates, PASS, PAES) 9

11 Objectives Regulatory Challenges Which Big Data? What are the Challenges 10

12 Challenges 90% Of the world s data has been created in the past 2 years. 24 months Frequency at which electronic healthcare data doubles 75%+ Percentage of patients expected to use digital health services in the future 11

13 Big Data Challenges Rate of accumulation 100,000 RCTs 424 million articles in 5600 journals >12 TB personal health data/lifetime Inaccessible, siloed data 80% of data unfindable Lack of interoperability Governance and privacy concerns Unstructured and heterogeneous 80% unstructured Valuable detail in unstructured text - Images, MRIs, X rays etc Multiple formats and provenance Uncertain quality Variable standards Lack of validation causality vs coincidence Managing bias and confounding 12

14 Regulatory Challenges Structured data (RCT) generated in accordance with strict guidelines and known provenance High certainty Unstructured, unvalidated data of unknown provenance more uncertainty 13

15 Interoperability and Harmonisation Common data models Minimal Data sets Standards Documenting the Strengths and Limitations Solutions Addressing privacy and Governance Robust validation 14

16 Objectives Regulatory Challenges Which Big Data? What are the Challenges New Regulatory Initiative 15

17 Chair: Thomas Senderovitz, DK Co Chair: Alison Cave, EMA Members: DE, DK, ES, FI, HU, IE, NL, NO, RO, UK 16

18 HMA/EMA Big Data Taskforce Characterisation and Mapping of Relevant Sources of Big Data Identification of Relevant Data Types Characterisation of Data Consider need for Data Standards Specific Regulatory Challenges External Survey 17

19 HMA/EMA Big Data Taskforce Define the usability and application of data Map regulatory decisions across product life cycle Where are the gaps/opportunities? Can Big Data plug the gap? Is data adequate? 18 What are the constraints in its use?

20 HMA/EMA Big Data Taskforce Describe the current status, future needs and challenges What is the current landscape in NCAs What datasets are currently used? What is the expertise in data analysis? What are the anticipated future needs? Generation of a set of recommendations and a roadmap 19

21 Objectives Regulatory Challenges Which Big Data? What are the Challenges New Regulatory Initiative Conclusions 20

22 Conclusions Vast amounts of healthcare data are continually being generated, offering huge opportunities but making it impossible to keep pace with all the information. Harnessing of the potential of big data by researchers and regulators is hindered by the fact that it is often unstructured, noisy and inaccessible. Deciding which data to collect starts by asking the right questions about the benefits sought and problems faced. Access to data is a significant hurdle especially for observational data. Mechanisms to integrate the data to generate meaningful knowledge is needed. Validation that associations are causal is key for data to support regulatory decision making. 21

23 ` Thank you European Medicines Agency 30 Churchill Place London E14 5EU

24 Workplan Characterisaton of Relevant Sources of Big Data and Defining Format Genomics Observational Data Social Media/M health Spontaneous ADRs Other omics Clinical Trial data IT/infrastructure Identify areas of usability and application of datasets Describe the current status, future needs and challenges Generate a list of recommendations and Big Data road map 23